if the upvotes/downvotes are part of a user feed… you could probably do some crunching on “repeated brigading posts”… but part of me is going… wouldn’t it be interested to make buckets for how similarly people’s opinions line up.

like… “subtract out the folks who just vote no on anything positive with MacOS”, e.g., the herd gets grouped and gets lesser power.

or… “rank a user that usually downvotes stuff in this group but suddenly upvoted it higher” (like neural exhaustion). So the more an account /just upvotes/ and it becomes tired due to the link between that user and those keywords.

You are welcome to contribute to the project here: https://github.com/LemmyNet

I know that the devs are only going to see feedback if it’s on the Github.